Question about credits

[MooMoo2]

Let's say User A finds a prime on LLRnet. Before he knows that his PC reported a prime, User B takes a look at the stats page, checks its +1 counterpart, and finds that it is a twin. The top siever and top LLR tester share the credit for the twin, but who else also gets it? User A, User B, or both of them?

[mdettweiler]

Originally Posted by MooMoo2

Let's say User A finds a prime on LLRnet. Before he knows that his PC reported a prime, User B takes a look at the stats page, checks its +1 counterpart, and finds that it is a twin. The top siever and top LLR tester share the credit for the twin, but who else also gets it? User A, User B, or both of them?

Hmm, tricky. I of course can't speak for Oddball, but my opinion would be that the finder's share of the credit should go entirely to User A. Technically it would indeed have been User A who found the -1 and User B who found the +1 and thus the twin, but since the latter is a much bigger find, it would be small consolation to User A to be simply noted in the server's log of primes. The gentlemanly thing to do in such a scenario would be for User B to voluntarily cede the full credit to User A. However, there's no guarantee that User B won't just happen to be an ungentlemanly person, so IMHO, there should be a rule that says the full credit goes to User A no matter what in order to discourage such a heist.

Again, though, this is just me talking as an ordinary member (not to confuse anyone: my green user title is from a different subforum ), so the final call would have to be Oddball's.

[Flatlander]

Just a thought.
If I found a muddy ring with a metal-detector and wasn't sure if it was brass or gold, then someone cleaned it off and found out it was gold before I got a chance, would it become half theirs?

[axn]

Originally Posted by Flatlander

Just a thought.
If I found a muddy ring with a metal-detector and wasn't sure if it was brass or gold, then someone cleaned it off and found out it was gold before I got a chance, would it become half theirs?

You can prove anything with a bad analogy

[Flatlander]

Originally Posted by axn

You can prove anything with a bad analogy

I thought it was quite good. lol.
It was finding the ring in the first place that took all the work.

[mdettweiler]

Indeed, while the analogy is not totally perfect (as, once you've initially claimed rights of salvage on the ring, it would be your property regardless of whether it's brass or gold and whether anyone cleans it), the basic sentiment does apply here. Even though primes don't really belong to anyone, User B's ceding full credit to User A is nonetheless the gentlemanly thing to do (since the whole point of prime crediting is to recognize the person(s) who put in the most work to find it).

[Oddball]

Originally Posted by mdettweiler

IMHO, there should be a rule that says the full credit goes to User A no matter what in order to discourage such a heist.

OK, here's the rule:

User A gets full credit unless one of these two things happen. Both of these cases are unlikely, but I'll mention them anyway:

1.) Incorrect residues for the +1 side. For example, 21399*2^480331-1 is prime, but the +1 side has a residue of F998107FB7E57082. If User B finds that the +1 residue is wrong, and that the +1 side is indeed prime, then he/she would share credit with the -1 user for the twin.

2.) The +1 side remains unchecked for ten days (240 hours) or more. For example, suppose xxx*2^yyy-1 is prime and nobody checks the +1 side for twins within 10 days. If User B finds that xxx*2^yyy+/-1 is twin after that, then he/she will also get some credit for the twin.

[mdettweiler]

Originally Posted by Oddball

OK, here's the rule:

User A gets full credit unless one of these two things happen. Both of these cases are unlikely, but I'll mention them anyway:

1.) Incorrect residues for the +1 side. For example, 21399*2^480331-1 is prime, but the +1 side has a residue of F998107FB7E57082. If User B finds that the +1 residue is wrong, and that the +1 side is indeed prime, then he/she would share credit with the -1 user for the twin.

2.) The +1 side remains unchecked for a week or more. For example, suppose xxx*2^yyy-1 is prime and nobody checks the +1 side for twins within a week. If User B finds that xxx*2^yyy+/-1 is twin after that, then he/she will also get some credit for the twin.

Sounds fair to me. However, IMHO the "statute of limitations", so to speak, on #2 should be a little longer. Since we have no email notification of primes on LLRnet, it is entirely possible that someone might never even notice that they found a prime--and you might not notice it either if you didn't happen to check the servers' status page sometime that day while the prime was displayed on the main page. (Sure, it's there in the full log, but it's easy to miss that a new prime showed up when quickly glancing there...especially once you get enough that they scroll off the screen.) In such a scenario, a prime could quite easily go an entire week without anyone knowing it was found--and if someone else spotted it, they could quite easily come in and "claim" the prime for themselves, and nothing could be done about it since, after all, they would have followed the rules exactly.

What might be better is to make it 2 weeks for now, while we don't have an email notification system for LLRnet. In the future, if we did get such a system, we could change it to 1 week. (Or, the server machine could check the +1 side automatically before it sends the notification email...we do leave one core idle for server stuff, so a couple of CPU-minutes here and there could easily be spared for such things.)

[Oddball]

Originally Posted by mdettweiler

IMHO the "statute of limitations", so to speak, on #2 should be a little longer.
...
What might be better is to make it 2 weeks for now, while we don't have an email notification system for LLRnet.

One week may be too short, but 2 weeks is too long. I've increased it to ten days, which is a reasonable amount of time.

[Oddball]

Just one note: a few people have asked me how credit will be distributed if one person is both the top siever and the top LLR tester. The rule is that runners-up don't get credit, so he will only share credit with the twin discoverer (along with TPS and the programs used in the project). For example, if user X discovers the twin and is the top siever and top LLR tester, he/she won't share credit with any other individuals.